Skip to content

Fix dk/dv autograd error on TPU flash attention #8685

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
qihqi merged 1 commit into from
Feb 7, 2025
Merged

Fix dk/dv autograd error on TPU flash attention #8685

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion torch_xla/experimental/custom_kernel.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -449,7 +449,7 @@ def fa_custom_backward(
if require_grad_ab:
grad_ab = grads[1]

if require_grad_k or require_grad_k:
if require_grad_k or require_grad_v:
payload, _ = trace_pallas(
_flash_attention_bwd_dkv,
q,
Expand Down